The proof is in the pudding
Strategist, Project Manager, User Experience Expert
Identify your base assumptions about your customers and their goals.
Be clear about what you are listening for. Make sure you set your standard of proof.
Beware of confirmation bias. Be honest with yourself about the data you collect.
Lesson: Creating Customer Personas with Julie Hamwood
Step #7 Testing: The proof is in the pudding
We've identified what the base assumptions are that we have about our customers and their problems. We really want to understand if our assumptions hold, are supported or not. There are a lot of different approaches, such as concierge services or running interviews or creating landing pages or mockups and having calls to action and seeing what the click-through rates are, going to the places where you think people are behaving in certain ways and seeing what trails they leave behind. There are a lot of different ways that you can identify what people's actual needs are.
Early in the process it's a really good idea to use interviews. At that point, you're still trying to just get a sense of what the shape of the landscape is. By going and talking with people and asking really open-ended questions that are still structured and you're very clear about what you're listening for, we're wanting to find out from people what it is that they need, because as entrepreneurs we have the capacity to meet a need. But we need to understand their need in order to align what we can make with what they need.
Let's assume we're going to use interviews. Before we actually even do the interview process, really make it explicit what your assumption is and structure it as a falsifiable hypothesis. So that means something you can actually collect data around, and when you collect that data, it's going to give you a clear signal -- Yes or No. Your hypothesis is supported or it isn't.
So you want to make sure the hypothesis is falsifiable. You also want to ensure that you've set the level at which you agree that it's supported or not supported. So if you're going to go out and do 5 interviews or 30 interviews, whatever the number is that you decide to do, what percentage of those interviews does the interviewee need to of their own volition, not being led, but of their own volition make statements that agree and support your hypothesis. It's really important to do this upfront, because we all get really wedded to our work and we all will have confirmation bias and we will look for evidence to support our point of view. So before we even start designing questions, we need to get clear about what our standard of proof is.
Confirmation bias is basically when you already have a point of view and you gather data and you basically select the data that confirms what your view is, and you have ways of minimizing and disregarding any data that speaks against that. We're just tuning into the thing that's most salient in our minds, and we have to protect ourselves against it if we want to be good businesspeople.